As companies begin to rely more and more on artificial intelligence (AI) technologies, the ethical questions involved have been becoming increasingly pertinent. Can AI or automated bots trained on data collected from human workers partially or fully replace the work that these people do? Should those people be compensated in any way if their data is used to create and train these bots? These questions have recently become a focus of study of a research team from both the Massachusetts Institute of Technology (MIT) and Stanford University.
In 2020, the researchers began an experiment testing OpenAI’s text-generation technology for use in customer service agents. The team of 5,000 agents, mostly from the Philippines, were given AI chatbot assistants generated from the data of top-performing workers. After analysis of the results, the team found that the AI tool had increased the productivity of their customer service team by 14%.
This has prompted researchers to ask what Li calls ‘the question of the century’- should top-performing employees be compensated if their data helps to create an AI system that can do their job? As the national Bureau for Economic Research argues, this could have a positive effect on employees that can help to incentivize them to continue to optimize their work and increase productivity.
Danielle Li, an economist at MIT’s Sloan School of Management, leads the research team and is examining the potential legal and ethical issues of data mining and using AI technology to replace human workers. Erik Brynjolfsson, director of the Stanford Digital Economy Lab, also supports the idea of rewarding workers whose data is used to create AI systems. He believes that compensating those workers would be beneficial to the business as they would have greater incentive to keep providing data that can help to boost productivity.
In related news, there has already been legal action taken against AI companies who use data of coders and artists for their projects. Reddit and Stack Overflow are also considering charging companies for access to their conversational data.
This research illustrates how generative AI tools can raise ethical issues and tensions, even within companies utilising them. Going forward, it will be interesting to see how the law may adapt to the new age of AI and how companies choose to address this issue of fair compensation for the data their employee’s generate.
